Why assume UG?
This paper deliberates for a number of linguistic features whether they are part of UG, i.e., specific to human <i>language</i>, or whether they are adapted from other cognitive capacities which were evolutionarily prior to language. Among others, it is argued that the distinction between predication and reference already belongs to the conceptual system, whereas the distinction between verb and noun (which is not identical with the former one) is one of the innovations of UG. It is furthermore argued that syntax in the sense that it deals with displacement (‘movement’) is a property of human language that lies outside of UG. The paper then discusses whether linguistic typology can contribute to our knowledge of UG, and whether aiming at this is a reasonable goal for typological research. It stands against Newmeyer’s position (this special issue) that typological evidence is essentially irrelevant for the construction of UG, as well as against Haspelmath’s position (this special issue), who argues that typological research can do without a concept of UG.